app.py

import gradio as gr
import requests
import io
import random
import os
import time
from PIL import Image
from deep_translator import GoogleTranslator

# Project by Nymbo

API_URL = "https://api-inference.huggingface.co/models/black-forest-labs/FLUX.1-dev"
API_TOKEN = os.getenv("HF_READ_TOKEN")
headers = {"Authorization": f"Bearer {API_TOKEN}"}
timeout = 100

def convert_to_png(image):
    """Convert any image format to true PNG format"""
    png_buffer = io.BytesIO()
    if image.mode == 'RGBA':
        # If image has alpha channel, save as PNG with transparency
        image.save(png_buffer, format='PNG', optimize=True)
    else:
        # Convert to RGB first if not in RGB/RGBA mode
        if image.mode != 'RGB':
            image = image.convert('RGB')
        image.save(png_buffer, format='PNG', optimize=True)
    png_buffer.seek(0)
    return Image.open(png_buffer)

def query(prompt, is_negative=False, steps=35, cfg_scale=7, sampler="DPM++ 2M Karras", 
          seed=-1, strength=0.7, width=1024, height=1024):
    if not prompt:
        return None

    key = random.randint(0, 999)
    API_TOKEN = random.choice([os.getenv("HF_READ_TOKEN")])
    headers = {"Authorization": f"Bearer {API_TOKEN}"}
    
    # Translate prompt
    try:
        prompt = GoogleTranslator(source='id', target='en').translate(prompt)
        print(f'\033[1mGeneration {key} translation:\033[0m {prompt}')
        prompt = f"{prompt} | ultra detail, ultra elaboration, ultra quality, perfect."
    except Exception as e:
        print(f"Translation error: {e}")
    
    print(f'\033[1mGeneration {key}:\033[0m {prompt}')
    
    payload = {
        "inputs": prompt,
        "is_negative": is_negative,
        "steps": steps,
        "cfg_scale": cfg_scale,
        "seed": seed if seed != -1 else random.randint(1, 1000000000),
        "strength": strength,
        "parameters": {"width": width, "height": height}
    }

    try:
        response = requests.post(API_URL, headers=headers, json=payload, timeout=timeout)
        response.raise_for_status()
        
        # Convert directly to PNG without intermediate format
        img = Image.open(io.BytesIO(response.content))
        png_img = convert_to_png(img)
        
        print(f'\033[1mGeneration {key} completed as PNG!\033[0m')
        return png_img
        
    except requests.exceptions.RequestException as e:
        print(f"API Error: {e}")
        if hasattr(e, 'response') and e.response:
            if e.response.status_code == 503:
                raise gr.Error("503: Model is loading, please try again later")
            raise gr.Error(f"{e.response.status_code}: {e.response.text}")
        raise gr.Error("Network error occurred")
    except Exception as e:
        print(f"Image processing error: {e}")
        raise gr.Error(f"Image processing failed: {str(e)}")

# Light theme CSS
css = """
#app-container {
    max-width: 800px;
    margin: 0 auto;
    padding: 20px;
    background: #ffffff;
}
#prompt-text-input, #negative-prompt-text-input {
    font-size: 14px;
    background: #f9f9f9;
}
#gallery {
    min-height: 512px;
    background: #ffffff;
    border: 1px solid #e0e0e0;
}
#gen-button {
    margin: 10px 0;
    background: #4CAF50;
    color: white;
}
.accordion {
    background: #f5f5f5;
    border: 1px solid #e0e0e0;
}
h1 {
    color: #333333;
}
"""

with gr.Blocks(theme=gr.themes.Default(primary_hue="green"), css=css) as app:
    gr.HTML("<center><h1>FLUX.1-Dev (PNG Output)</h1></center>")
    
    with gr.Column(elem_id="app-container"):
        with gr.Row():
            with gr.Column(elem_id="prompt-container"):
                with gr.Row():
                    text_prompt = gr.Textbox(
                        label="Prompt", 
                        placeholder="Masukkan prompt dalam Bahasa Indonesia", 
                        lines=2,
                        elem_id="prompt-text-input"
                    )
                
                with gr.Accordion("Advanced Settings", open=False):
                    negative_prompt = gr.Textbox(
                        label="Negative Prompt", 
                        value="(deformed, distorted, disfigured), poorly drawn, bad anatomy, wrong anatomy, extra limb, missing limb, floating limbs, (mutated hands and fingers), disconnected limbs, mutation, mutated, ugly, disgusting, blurry, amputation",
                        lines=3
                    )
                    with gr.Row():
                        width = gr.Slider(1024, label="Width", minimum=512, maximum=1536, step=64)
                        height = gr.Slider(1024, label="Height", minimum=512, maximum=1536, step=64)
                    with gr.Row():
                        steps = gr.Slider(35, label="Steps", minimum=10, maximum=100, step=1)
                        cfg = gr.Slider(7.0, label="CFG Scale", minimum=1.0, maximum=20.0, step=0.5)
                    with gr.Row():
                        strength = gr.Slider(0.7, label="Strength", minimum=0.1, maximum=1.0, step=0.01)
                        seed = gr.Number(-1, label="Seed (-1 for random)")
                    method = gr.Radio(
                        ["DPM++ 2M Karras", "DPM++ SDE Karras", "Euler", "Euler a", "Heun", "DDIM"],
                        value="DPM++ 2M Karras",
                        label="Sampling Method"
                    )

        generate_btn = gr.Button("Generate Image", variant="primary")
        
        with gr.Row():
            output_image = gr.Image(
                type="pil",
                label="Generated PNG Image",
                format="png",  # Explicitly set output format
                elem_id="gallery"
            )
        
        generate_btn.click(
            fn=query,
            inputs=[text_prompt, negative_prompt, steps, cfg, method, seed, strength, width, height],
            outputs=output_image
        )

app.launch(server_name="0.0.0.0", server_port=7860, share=True)